Analytical Data Platform:
AirMettle's Analytical Data Platform enables users to quickly obtain key insights from large data sets
(records, multi-dimensional data, images, video) by analyzing the data in the storage tier. Our patented technology
integrates massively parallel processing (including AI inferencing) of distributed data within a software-defined storage
framework—accelerating analytics by orders of magnitude. By processing data in place, the platform eliminates the
need to transfer large datasets to separate compute clusters for analysis, significantly reducing network traffic and latency,
as well as analytics compute, memory, and storage requirements—making real-time analytics practical even at petabyte scales.
For further information, the
AirMettle User Guide is available for
download, or contact
info@airmettle.com
Hierarchical Multi-Dimensional Histograms (HMDH):
AirMettle’s Hierarchical Multi-Dimensional Histograms solution enables high-performance exploratory analysis of large,
complex datasets. Our solution uses parallel processing of stored data to rapidly characterize floating-point and integer
data distributions in up to four dimensions in a single pass, efficiently stores the results, and provides and API for advanced
querying (e.g., range queries, percentile calculations, etc.). HDMH files are compact “fingerprints” containing statistical
properties of massive datasets—including correlations between parameters. These compact histograms are typically over 1,000
times smaller than the raw data, while retaining precise counts and addressing at each hierarchical level. Histogram files also
can be compared, facilitating the analysis of variations across datasets. HMDH has broad application in research, manufacturing,
finance, and other industries. For further information see
our post on LinkedIn
or contact
hmdh@airmettle.com
